Gun sound -- frequency analysis

[btb4198]

Small caliber guns have higher magnitude at higher frequencies and larger caliber guns have higher
magnitude at lower frequencies. is this true ? because I am not see this in my test:

that is the fft of a 9MM gun,
As one can see there are low magnitudes at low frequencies and higher magnitudes at higher frequencies.
Also the peck frequencies are around 807 Hz. A 9MM handgun is a Small caliber gun.

that is an fft of a larger caliber gun and the magnitudes are higher at high frequencies ...
am I missing something ?

 

[rbelli1]

Can you set you software to RMS and then integrate on both sides of what you might call "low frequency"? That will get you the total energy at "high" and "low" frequencies. You can compare these pairs for different types of guns.

Scale all of them so one side is 1. The other number will give the answer you seek.

BoB

 

[btb4198]

I am using audacity, but I am not sure what you mean by RMS

also is 804 not a low frequency ?

 

[btb4198]

I do no thing audacity has a RMS mode:
Root mean square right ?

 

[rbelli1]

The spectrum can be scaled as dB which is a relative measure and also as RMS (yes Root Mean Square) which is an absolute value. There should be a setting somewhere in your software. You may need a different plugin.

As to 804Hz being low or high frequency that depends on what you are doing. My human perception would put it at middle or high frequency. (YouTube 804Hz for a sample)

BoB

 

[btb4198]

ok there is this :

this is Linear frequency .
so what are the different between plotting it as Linear frequency vs Log frequency ?

 

[btb4198]

this is the 9MM one again :

but with Linear frequency
it looks the same as the other one

 

[rbelli1]

dB is a logarithmic scale so plotting both axes at the same scaling gives a better representation of the data. I don't know of the use of plotting X at a linear scale and Y and a logarithmic scale.

BoB

 

[btb4198]

ok but do you know why it is not working ?
do you know another program I should try ?

 

[rbelli1]

Your program is working just fine. There is probably a way of analyzing the data as dB to get the answer you want. I just select the parameters I need on my oscilloscope and use the cursors for the frequency range needed. I am not familiar with using PC software for this work.

I'm afraid that is about all I know about this subject. Hopefully someone with more knowledge will come along to help further.

BoB

 

[Baluncore]

It is not the bore of the gun that decides the spectrum. It is the position of the microphone and the exit velocity of the projectile.

Supersonic ammunition will generate a shockwave that is a step function with relatively little low frequency energy. Lower velocity ammunition, or a muzzle brake, will generate a broad low frequency spectrum typical of a trapezoidal waveform.

High frequencies are steeply attenuated in air. The distance to the microphone will be critical.

Any echo will generate a frequency dependent interference pattern. How do you eliminate echoes from the microphone. Do not forget echoes from the ground.